Text Mining for Criminal Investigation Offices

    1 June, 2006

    2006-2008. Government funded project to develop text and video mining tools for crime investigation. Collaboration with Tamás Szirányi's http://www.sztaki.hu/~sziranyi/ team for video event detection, modeling, classification and annotation.

    MONLINGV (Development of a knowledge management tool for the European criminal investigation offices)


    The relation between criminology and IT has developed a lot in the past years. Organisations investigating crimes usually have a vast amount of data. The available data have potential correlations where fact-finding can only be solved partly by advanced IT-solutions, professional police work and partly with the help of data mining and e-learning based tools classifying computer-based linguistic, textual and image information. With the help of such analyzing systems relations between people can be discovered and conclusions can be drawn from the perpetrators’ method, events and their geographic location. These analyses are important not only for the investigations but also for prevention.

    One big innovation of the project is that it attempts to develop practical text-processing algorithms (e.g. categorizing texts, search upon content) built upon the basic linguistic and knowledge presentation technology functionally. The unique goal of the project is the interpretation of events, images and videos in a cognitive way. The definition of the properties of the searched image content will be realised on the basis of search in text files and by this the content of the available images and videos can be interpreted and contents in question searched.


    Montana Inc., MTA SZTAKI, Hungarian National Police, Research Institute for Linguistics HAS, Montana GmbH (D), Renate Burian (D)